Here's one topic Americans can bank on hearing about in next week's State of the Union address: gun control. The reaction to President Obama's announced gun-control measures this week was swift and entirely as expected. Gun-control advocates and many Democrats applauded his efforts; gun-rights groups and many Republicans loudly denounced the orders as executive overreach.
Expanded background checks are central to the president's proposals. His order doesn't rewrite existing laws, but it would broaden the scope of who is in the gun-selling business. It would require more gun sellers online and at gun shows to be licensed (and perform checks) among other things.
"Let me be clear: It's not where you are located but what you are doing that determines whether you are engaged in the business of dealing in firearms," Attorney General Loretta Lynch told reporters this week.
So would those extra checks bring down America's high levels of gun deaths? Gun policy experts who spoke to NPR say it could, but if so, that it would only make a dent.
Here's a look at the evidence:
What research says
Two recent studies provide evidence that background checks can significantly curb gun violence. In one, researchers found that a 1995 Connecticut law requiring gun buyers to get permits (which themselves required background checks) was associated with a 40 percent decline in gun homicides and a 15 percent drop in suicides. Similarly, when researchers studied Missouri's 2007 repeal of its permit-to-purchase law, they found an associated increase in gun homicides by 23 percent, as well as a 16-percent increase in suicides.
Those are some huge results — one expert called the Missouri study "the strongest evidence that background checks really matter," as The New Republic reported — but as with lots of social-science research, there's some fuzziness as to what the results mean. One caveat is that these studies aren't about background checks alone. Instead, they're about permit-to-purchase laws, under which people had to go to local law enforcement to get a permit and, therefore, a background check.
That difference might have impacted the results, explained Daniel Webster, a co-author on both studies. He said that being forced to get a permit from law enforcement might do more to deter a straw purchaser, for example, than getting a check at a nearby store.
Furthermore, he added that because so many factors influence gun violence in different ways, it's hard to say how much the effects seen in Connecticut and Missouri would also happen in other states. In addition, a stand your ground law enacted in Missouri in 2007 may have affected the results.
Still, other academic research points to the laws' effectiveness as well. In a 2015 analysis of studies published over the course of 15 years, Webster and co-author Garen Wintemute found that expanding background checks could "have protective effects against lethal violence," and that permit-to-purchase laws in particular help curb murders and suicides.
They also found that background checks help keep guns out of the hands of criminals, but that it's less certain whether that in turn leads to less violence.
There's no perfect consensus on how well background-check laws work. A 2000 study found that the 1994 Brady Act — which instituted not only background checks but waiting periods at first — did not reduce either homicide or suicide rates.
A CDC task force also found in a 2003 review "inconsistent findings" as to whether restricting gun access through background checks works and insufficient evidence as to whether an array of other gun laws are effective. However, the CDC also said that its findings didn't mean that gun laws don't work; rather, it said it needed to study the topic more.
Gun-policy researchers say they want to better study background checks (as well as many other policies), but a couple of hurdles stand in the way. Part of the problem is that good studies on the effectiveness of background checks are pretty rare, according to Webster. One reason is that it's hard to find good test cases to study.
"There's not a lot of change or variation [in laws] to study in recent times," he said. "The vast majority of these laws have been on the books for many, many decades."
Another expert blamed the federal government.
"One of the big problems is that the feds have not funded good research in this area," said David Hemenway, director of the Harvard Injury Control Research Center and an expert on firearm-related injuries.
He points to federal restrictions, passed in 1996, that said the Centers for Disease Control could not use its funding to "advocate or promote gun control." That caused the CDC to back away from gun research almost entirely.
Outside organizations could pick up that slack, Hemenway added, but they have not done so. "The foundations haven't done a good job, because it's such a controversial area," he said. You don't want to get involved. So we know some things, but we don't know as well as you would hope, given the enormity of the problem."
What recent shootings tell us
While some scholarly evidence suggests that background checks reduce crime, seeing evidence in recent mass shootings is tougher. As the New York Times found in a December investigation, the guns used in many recent high-profile shootings were purchased legally by people who passed background checks.
Importantly, though, to the extent that background-check laws on the books might have prevented mass shootings, it's impossible to compile similar lists of incidents that would have occurred, were it not for those laws.
One other thing recent shootings say is that the current background-check system has some gaping holes in it. For example, FBI Director James Comey said in July 2015 that Dylann Roof, who is accused of killing nine at a South Carolina church last year, should not have passed a background check. Because information about his admission to a narcotics charge never reached an FBI examiner handling his check, as the Washington Post reported, Roof was able to buy his gun.
In addition, some states are doing a poor job of submitting mental health records to NICS, as Politico's Kevin Cirilli writes, allowing some sick people to obtain guns. Cirilli points to Virginia Tech shooter Seung-Hui Cho, who had a history of mental illness before he killed 32 people in 2007.
As it stands, around 1.6 percent of 148 million background checks (that is, more than 2 million) between 1994 and 2012 were denied, according to federal statistics.
What the statistics say
One of the most important questions to this discussion is impossible to answer precisely: how many guns are obtained without background checks? While there aren't exact numbers on this, the figure could still be substantial. Using 2004 data, around 18 percent of gun transactions involved private sellers, buyers' family members or friends or "other" sources, as the Washington Post's Glenn Kessler found last year. A majority of those sources were not licensed dealers (and therefore were not required to conduct background checks).
According to the figures cited by Kessler, 7 percent of guns were obtained from gun shows (and many of those sales probably underwent background checks).
But data suggests that gun shows don't directly supply many of the guns used in crimes. Spokespeople from the National Rifle Association and National Shooting Sports Foundation, a trade group for gun sellers, both also pointed NPR to government data showing that less than 1 percent of prison inmates in 1997 said they got their guns from gun shows. Meanwhile, nearly 80 percent obtained their guns from friends, family or "street" (illegal) sources.
All of this very well may mean that, as gun-rights advocates like Florida Republican Sen. Marco Rubio often point out, criminals will simply obtain guns through some avenue other than stores. That would mean that background checks don't deter those people, and, therefore, that expanding them to more online or private or gun show sales would do little.
But there are other possible conclusions. A recent study of offenders in the Chicago area found most obtained their guns from "personal connections, not from gun stores or by theft." While that study suggested to some that background checks are ineffective, one of the authors, Duke University's Philip Cook, disagrees.
"This research demonstrates that current federal and local regulations are having a big effect on the availability of guns to criminals in Chicago," he said in a release. "They can't buy their guns from stores, the way most people do, and are instead largely constrained to making private deals with acquaintances, who may or may not be willing and able to provide what they want."
Lawrence Keane, general counsel for the National Shooting Sports Foundation, told NPR that "the industry has always been supportive of the background check system," though he also said he has doubts about how much good the new proposals will do.
Researchers Hemenway and Webster both think the president's executive actions could have a modest effect on gun violence. For his part, Hemenway thinks universal background checks would be an effective first step, but what he thinks would be more fruitful in the long term has more to do with innovation than legislative action.
"In the long run, we should be spending a lot of money on figuring out technological fixes," he said. "The easiest one is to make guns better for home protection and much, much less dangerous and less likely to be stolen."